Unravelling the effect of data augmentation transformations in polyp segmentation.

Luisa F Sánchez-Peralta Artzai Picon Francisco M Sánchez-Margallo José Blas Pagador

Published in: International journal of computer assisted radiology and surgery (2020)

Despite being infrequently used, pixel-based transformations show a great potential to improve polyp segmentation in CVC-EndoSceneStill. On the other hand, image-based transformations are more suitable for Kvasir-SEG. Problem-based transformations behave similarly in both datasets. Polyp area, brightness and contrast of the dataset have an influence on these differences.

Keyphrases

deep learning
convolutional neural network
magnetic resonance
big data
artificial intelligence
rna seq
risk assessment
contrast enhanced